openMSX 0.10.1 released

openMSX 0.10.1 released

by Manuel on 01-05-2014, 23:26
Topic: Emulation
Tags: Catapult, openMSX
Languages:

The openMSX Team is happy to announce a new release of openMSX and its GUI launcher Catapult. This openMSX release fixes mostly bugs we introduced with 0.10.0 or were already present for a bit longer. There are also some small but nice additions.

Here's the list with all highlights:

MSX device support

  • fixed bug in YM2413 (FMPAC), that was audible in BPS Tetris
  • added Sharp/Epcom HB-4000 80 column cartridge extension

New or improved emulator features

  • extended save_msx_screen script to take VDP(24) into acount
  • fixed crash on Visual Studio build when using wav files (e.g. in the
  • cassetteplayer)
  • fixed crash on 32-bit Visual Studio build when recording videos without the -doublesize option
  • fixed crash on 32-bit Visual Studio build when using scale_factor 3 and SDL renderer
  • fixed waves shown upside-down in the SCC Viewer OSD widget
  • fixed crash when loading a savestate that has JoyTap plugged in
  • performance improvements
  • fixed handling of corrupt hardware configuration XML file
  • fixed crash when MSX with TC8566AF FDC (e.g. turboR) crashed
  • fixed possible hang up when reversing with harddisks
  • fixed Tcl error when changing horizontal stretch in OSD menu
  • fixed displaying of error messages on OSD
  • Better Mac OS X MIDI support
  • fixed issues with on-screen-keyboard on Android 4.4
  • added default keybindings for controllers targeted at gaming
  • added mapping of B-control button to press CTRL or SHIFT in OSD keyboard
  • added fine grained control in which slots carts and extensions will end up
  • added support for single sided disk drives for dir-as-disk
  • added support for hard disk images to the OSD menu
  • added warning if ROM images with larger size than supported on real Konami and Konami SCC mappers are used

Build system, packaging, documentation

  • updated Windows build to Visual Studio 2013
  • made compilation faster by removing more than 250 unnecessary includes
  • even more use of C++11 features
  • upgraded Tcl to 8.5.15
  • added build support for 64-bit ARM
  • added 256x256 pixel openMSX logo image and use it e.g. for Mac OS X icon
  • fixed staticbindist build with more recent clang releases and for Mac OS 10.9 SDK
  • do not link against libraries in /usr/lib anymore on Mac OS X, only link
  • against frameworks. This enables to compile against the default SDK.

Also new in this release is a significant update to Catapult. Apart from a big cleanup and some bug fixes, some new features were added. Most notable is support for hard disks and support for changing cartridges at run time.

Catapult changelog

  • fixed disk image file not shown on UI when inserting a disk twice
  • made behaviour the same for all media (was inconsistent at places)
  • added support for changing cartridges at run time
  • add DESTDIR support (for packagers)
  • made Catapult code compatible with wxWidgets 3.0. It requires 2.8 or 3.0 now.
  • fixed several bugs on the session page
  • in the history of previously used media, now also store the accompanying IPS patches
  • fixed setting the video source when there's a space in its name (e.g. Neos MA-20)
  • removed annoying CTRL-A short cut for the About dialog
  • fixed layout of About dialog
  • added support for hard disk images (works of course only when such an extension is present)
  • fixed Catapult interrupting a replay because it plugged in the cassette player
  • fixed fps-display in the Catapult status bar
  • added build support for 64-bit ARM
  • added higher resolution Windows icon and logo in about box

Happy MSXing!

Relevant link: openMSX website

Comments (69)

By gdx

Prophet (2932)

gdx's picture

02-05-2014, 02:20

Can you add support for MegaFlashROM of ASCII type (8/16K) in a next version ?

By Meits

Scribe (5461)

Meits's picture

02-05-2014, 07:21

Good job guys...

By Manuel

Ascended (15631)

Manuel's picture

02-05-2014, 08:43

gdx - I've got information and some documents about it, but I can't promise anything. I do think it's worth implementing it.

By Prodatron

Paragon (1788)

Prodatron's picture

02-05-2014, 19:37

Impressive changelog!

By giuseve

Paladin (733)

giuseve's picture

03-05-2014, 10:48

I'm having little problem with LEFT ARROW key.
It doesn't work.
I know I must use BIND, but I don't know all the names of keyboard button.
I mean: BIND LEFT_ARROW lEFT should work?
I've looked to the command line guide, but i did't find any kind of keyboard map.

Wonderful job anyway !!!
I hope one day catapult (orsomethinglikethis) will have alabel for machine customization, a label for diskmanipulator tool.

Regards

By Manuel

Ascended (15631)

Manuel's picture

03-05-2014, 20:18

You don't need to bind the left arrow key. It is mapped to the left cursor key by default.

Does it never work? Or only with certain software? Did it work correctly before?

By giuseve

Paladin (733)

giuseve's picture

03-05-2014, 22:08

I know it is mapped by default.
But since many days it does not work anymore.
I thought that installing ver.10.1 it would working back.
I was wrong.
It happens always, i do not know why.

By Manuel

Ascended (15631)

Manuel's picture

03-05-2014, 22:15

I also don't know, I've never heard anyone complain about this before. Why didn't you report it earlier?

You can try some older openMSX versions to see if there is a version which the problem did not occur. All old versions are still available from https://sourceforge.net/projects/openmsx/files/openmsx/

By wouter_

Champion (415)

wouter_'s picture

04-05-2014, 10:41

@giuseve: Does it help if you delete your settings.xml file? OpenMSX will recreate it with all settings reverted to the default.
But please make a backup first so that, if this turned out to be the culprit, we can compare both versions and prevent this bug(?) from happing to other users.

By ren

Paragon (1217)

ren's picture

04-05-2014, 14:42

Cool! Cool
Any reason the SDLGL-FB renderers aren't included in the official releases?
I actually prefer to use SDLGL-FB32. It seems to give the same performance as -PP, but I like the blurring better in that mode.. (for me -PP blur 1 is already way too blurred... I guess it enables bilinear filtering or something alike when blur is activated?)

A while ago I was playing around with Powerstrip and open- and blueMSX to get the best vsync results.. I actually discovered my monitor can handle 50Hz (although the specs say 56Hz). Combining the right monitor freq with vsync enabled in my GPU driver options, and disabling desktop composition (Win7) did the trick for me - total smoothness and a happy man :-)
(I hope I'm not ruining my monitor this way btw, but being able to enjoy total smoothness on 50Hz productions is worth it I guess?? ;-))

Do you know what frequencies (openMSX) PAL & NTSC precisley operate at? And does this differ between devices, or is this a standardized value?
I searched for this, but couldn't really find a decisive answer.

For PAL I seem to get the best results with a freq of around 50,142 Hz. For NTSC my default monitor freq of 59,883 Hz seems to suffice.

Thanks!

By Manuel

Ascended (15631)

Manuel's picture

04-05-2014, 16:46

The frequency is: (6 * 3579545) / (1368.0 * (([vdpreg 9] & 2) ? 313 : 262)))

PAL: 50.15897371223586 Hz
NTSC: 59.92274340431231 Hz

So, the framerate openMSX generates is always one of these 2 (if frameskipping is disabled of course).

By ren

Paragon (1217)

ren's picture

04-05-2014, 15:57

Thanks Manuel, helpful indeed. Any comments on my other questions? Smile

By Manuel

Ascended (15631)

Manuel's picture

04-05-2014, 17:05

I don't know exactly about the FB renderers. IIRC mth added them at some point and I forgot the story about them, so I asked him to reply here.

Oh, they were added by Wouter in 2005 Smile

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

05-05-2014, 10:08

@Manual,
I am gonna try openMSX, read all the features on the website, looks very impressive Smile
Mainly going to use it to compile MSX-C and test on several system setups.

How much is the speed up percentage maximally possible by the way?

By wouter_

Champion (415)

wouter_'s picture

05-05-2014, 19:14

Wolverine_nl wrote:

How much is the speed up percentage maximally possible by the way?

Do you mean how much faster than real-time can openMSX emulate? This of course heavily depends on the type of software you emulate and on the speed of your host computer. But to give you some rough ideas: most PC hardware should have no problem running openMSX at 10x real-time speed.

Most often video rendering is the bottleneck, and occasionally sound emulation takes a significant part. So if you allow to skip rendering some frames (maxframeskip setting) and if the MSX software doesn't produce sound, it's possible to achieve 50x-100x real-time speed.

For certain MSX software even higher speeds are possible. E.g. the ZEXALL test almost exclusively needs Z80 emulation (there's no sound and only occasionally it prints something to screen). On a real MSX machine this test takes ~4.6 hours to complete. I just retested on openMSX .. with some setting-tuning it finishes in less than 1 minute on my system (that's ~275x real-time speed, or equivalent to a Z80 running at ~1GHz)!

I would guess that C-compilation on MSX is also mostly Z80 emulation (and a bit of disk access), so you may be able to achieve similar speeds.

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

05-05-2014, 19:55

@ Wouter:
Yes it is pure for c-compilation and other processes, testing stuff in realtime also in a later stadium and also on a real MSX.
Going to install it now. Smile

By Paulo Volkova

Expert (104)

Paulo Volkova's picture

05-05-2014, 20:54

Hello.

I installed "OpenMSX 0.10.1-26" but this version don't have Catapult. This is normal?

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

05-05-2014, 21:15

I have installed 0.10.1 and I am using Catapult Tongue
Still busy with connection to harddisk image, not really working, will look for answer first Wink

By Manuel

Ascended (15631)

Manuel's picture

05-05-2014, 21:20

Paulo - Just install the official release. (But all Windows builds have Catapult.)

By AxelF

Champion (340)

AxelF's picture

05-05-2014, 21:33

Great job, this emulator keep getting better and better Smile

I noticed 2 minor things in catapult
-1- If i mount a ROM file in catapult while the emulator is running it does not insert it.(in console it does work )
-2- Using the scroll wheel of the mouse on all the pulldown boxes in catapult, only toggles thru the last 2 entry's
would be nice to go way back in history loaded files.

Keep up the great work guys Wink

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

05-05-2014, 21:41

Works great with my compiling in MSX-C, very solid performance, also testing my game engine at 3x screensize is very helpfull, going to use this great emulator! Cool

Only had 1 thing, it seems you need to let the emulator create a IDE HD image first before you can select your own, got errors before i let the emulator run its course, is there a reason for that order?

By Manuel

Ascended (15631)

Manuel's picture

06-05-2014, 17:17

Thanks for all the feedback!

@AxelF: damn, you're right... It works fine when selecting from a combobox, but browsing a new file doesn't work... I can't believe this slipped through our extensive tests. Thanks for reporting.
I can also reproduce the scroll wheel issue. I'm going to investigate with the wxWidgets toolkit folks if this is a bug on our side or a limitation of the toolkit.

@Wolverine_nl: I'm not sure what you mean here. I created a copy of my ide extension, selected it in Catapult and selected a harddisk image in Catapult. openMSX started up with it just fine and no default hd.dsk was created.
So, what kind of errors do you mean exactly? What exactly did you try?

By Manuel

Ascended (15631)

Manuel's picture

06-05-2014, 22:07

ren, about SDLGL-FB32 - what is the added value for you compared to the plain SDL renderer?

By ren

Paragon (1217)

ren's picture

06-05-2014, 23:34

@manuel as I said; I like the blurring better.. Smile
(alas, no glow effect, but I can live with that.. Smile)

Ah wait - you're talking about the plain SDL renderer..
Well, with that one I don't get a nice/proper (smooth) vsync (and the scaling is different I guess? I have to check again..) (tomorrow Wink)

By ren

Paragon (1217)

ren's picture

07-05-2014, 09:57

Brand new day! (more MSXing & other nerd-stuff to do! ;-))

So, it has nothing to do with the scaling, but with plain SDL stuff like scrolls are really choppy. SDLGL-FB32 works like -PP in that regard; even without vsync enabled, it's quite smooth (depending on the relation MSX freq <> monitor freq of course) (with a little hiccup at a regular interval).

Btw; with regular SDL & -FB32 I can't set scaling to 4x (it will fall back to 3x then)? I *can* set it to 4x with -PP.

By Samor

Paragon (2027)

Samor's picture

07-05-2014, 15:30

nice to see that old slingshot finally got an update Hannibal

By Manuel

Ascended (15631)

Manuel's picture

09-05-2014, 10:08

Paulo Volkova wrote:

I installed "OpenMSX 0.10.1-26" but this version don't have Catapult. This is normal?

No, but this has been fixed now. Thanks for trying development builds!

By ren

Paragon (1217)

ren's picture

09-05-2014, 11:27

Hey, I am trying development builds!
(just joking.. Wink)

By Manuel

Ascended (15631)

Manuel's picture

09-05-2014, 12:06

Well, the more people try development builds and report the problems they have with them, the less bugs we have in the release! So, I appreciate it a lot when people try development builds and report issues. Keep them coming, all!

By Manuel

Ascended (15631)

Manuel's picture

10-05-2014, 22:32

Manuel wrote:

@Wolverine_nl: I'm not sure what you mean here. I created a copy of my ide extension, selected it in Catapult and selected a harddisk image in Catapult. openMSX started up with it just fine and no default hd.dsk was created.
So, what kind of errors do you mean exactly? What exactly did you try?

Wolverine_nl??

Quote:

-1- If i mount a ROM file in catapult while the emulator is running it does not insert it.(in console it does work )

Can you please test with the latest development build? It should be fixed now.

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

13-05-2014, 10:47

Hi Manuel,
Sorry for the late reply, was quite busy actually using openMSX Tongue
I installed the previous build again on another laptop a few days ago and I tried to repeat the handlings I did, to replicate the issue, but it didn't occur, so it was or either a config thing I did wrong in the xml of my MSX hardwareconfig, or something else not quite sure. The first time I deleted my config and found one online and used that one and that did it.
I also wanted to know why a hard-disk image of 100mb was automatically being made in the first place when no image is selected, read that somewhere and actually saw the image appear while fooling around with Catapult, maybe better to put an extra Tab in the Catapult shell to create one manually or do a pop-up asking this to the user, just an idea Wink

I will try the new build later today before my c coding session as well.

cheers!

By Manuel

Ascended (15631)

Manuel's picture

13-05-2014, 17:29

Wolverine_nl wrote:

Hi Manuel,
Sorry for the late reply, was quite busy actually using openMSX Tongue

That's good, I guess Smile

Quote:

I installed the previous build again on another laptop a few days ago and I tried to repeat the handlings I did, to replicate the issue, but it didn't occur, so it was or either a config thing I did wrong in the xml of my MSX hardwareconfig, or something else not quite sure.

Some questions:
1) did you try with a copy of the IDE extension? So copy ide.xml to ide2.xml for instance. Because once you have the autogenerated harddisk image, the test isn't the same anymore, of course.
2) What were you doing with configs exactly and why?

Quote:

The first time I deleted my config and found one online and used that one and that did it.

Deleted what config and what did you find online? And why?

Quote:

I also wanted to know why a hard-disk image of 100mb was automatically being made in the first place when no image is selected, read that somewhere and actually saw the image appear while fooling around with Catapult, maybe better to put an extra Tab in the Catapult shell to create one manually or do a pop-up asking this to the user, just an idea Wink

The idea is that you can use the IDE extension directly. So when you use it without specifying a harddisk image, it autocreates one if it doesn't exist and uses that default one. Can you explain why this behaviour would be wrong/annoying/inadequate? (Except that it costs you 100MB disk space, but that's not too much these days, is it? Wink)

Quote:

I will try the new build later today before my c coding session as well.

Note that we didn't change anything regarding harddisks in that new build. The biggest change you'll notice now is that Catapult issue that AxelF found. But I hope AxelF will confirm that one.

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

14-05-2014, 09:34

Yes, i needed to elaborate it more it seems -=lol=-
the config was for the machine a was emulating, my own machine HB-F700 Wink was new to OpenMSX, now I know what does what.

I do have minor thing I encountered yesterday when compiling (my own project) and testing and using resets and stop/start allot. It seems inconsistent, the time it takes for the screen to popup and start the emulator, it is sometimes seconds in difference, So i use reset more.

Thats all for now, compiling on msx has never been faster Smile

By Manuel

Ascended (15631)

Manuel's picture

14-05-2014, 17:13

There is a Sony HB-F700P and HB-F700D coming with openMSX... so I'm still not sure what you mean...?

And can you still try to check on the harddisk stuff? See my questions 1 and 2 above.

Start/Stop means a full startup of openMSX. I'd expect it to be faster when you do it more often.

By Randam

Paladin (916)

Randam's picture

14-05-2014, 20:20

The link to nekolauncher on the openmsx site is no longer working. IT moved to:
https://nekocanmac.wordpress.com/software/

By Manuel

Ascended (15631)

Manuel's picture

14-05-2014, 20:49

Thanks updated.

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

15-05-2014, 11:09

Quote:

There is a Sony HB-F700P and HB-F700D coming with openMSX... so I'm still not sure what you mean...?

I've put the system roms in the userfolder, but it didn't recognize the roms in the first place, couldnt select the HBf700p from the list in Catapult, only C-bios setups. So i put an xml file in share/machines/ then it made it possible to select my machine.
I did saw the xml files of the pre installed machine, but it didn't show them after installation in the pulldownlist in Catapult.

Quote:

And can you still try to check on the harddisk stuff? See my questions 1 and 2 above.

I will try this probably tomorrow.

Quote:

Start/Stop means a full startup of openMSX. I'd expect it to be faster when you do it more often.

Clear as water, but will pay even more attention to this now Smile

By Manuel

Ascended (15631)

Manuel's picture

15-05-2014, 12:59

Wolverine_nl wrote:
Quote:

There is a Sony HB-F700P and HB-F700D coming with openMSX... so I'm still not sure what you mean...?

I've put the system roms in the userfolder, but it didn't recognize the roms in the first place, couldnt select the HBf700p from the list in Catapult, only C-bios setups. So i put an xml file in share/machines/ then it made it possible to select my machine.
I did saw the xml files of the pre installed machine, but it didn't show them after installation in the pulldownlist in Catapult.

Catapult doesn't show machines if they don't work (i.e. no ROMs are available). There's no point in moving XML files around, just make sure you have the correct systemroms in the share/systemroms dir, then rescan with Catapult and it will show them. The systemroms need to be exactly right, openMSX checks the sha1sum of them. Look here for some interesting zip file: http://www.msxarchive.nl/pub/msx/emulator/openMSX/

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

15-05-2014, 13:31

Ok, thanks for the link Smile

By Manuel

Ascended (15631)

Manuel's picture

16-05-2014, 09:56

This works 100% (and is explained in the manual):
1) install with installer
2) put the systemroms in share/systemroms

and everything runs 100% guaranteed. There is no need to do anything else. Really.

Really.

Really!

By flyguille

Prophet (3029)

flyguille's picture

24-05-2014, 00:51

OpenMSX is too dangerous!...

I just was working on my stuff, ans when get back to the openMSX window it show the "time bar" shit, and accidentally it get back in time and ruined all that I was doing.... this shitty timebar grrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr, I never be capable on forwarding it again LOST everything.

By Manuel

Ascended (15631)

Manuel's picture

24-05-2014, 12:06

That sucks! In such cases, where you don't want/need the reverse bar functionality, you can simply disable it: in the console type toggle_reversebar; you can change the default behaviour with this setting: http://openmsx.sourceforge.net/manual/commands.html#auto_ena...
You can also let openMSX automatically save a replay (i.e. all your input so that you can restore the reverse history by loading it), but that is a bit of hacky script. If you're interested I can explain how it works.

By sd_snatcher

Prophet (3032)

sd_snatcher's picture

24-05-2014, 15:49

For me, the reverse feature is essencial, as it allows the step_back feature when debugging. I can't live without it anymore. Smile

But I agree that the reverse bar is dangerous for programmers. Long ago I disabled it to avoid accidents.

Manuel, I tried the "set auto_enable_reverse on" feature, but it doesn't keep that setting after I close and reopen openMSX... I'm typing the following commands on the console:

set auto_enable_reverse on
save_settings

->The reverse feature is still disabled.

I'm running openMSX 0.10.1 on Mac OS-X 10.9.3.

By Manuel

Ascended (15631)

Manuel's picture

24-05-2014, 16:04

How did you disable reverse in the first place then? It's on the 'gui' setting by default.

EDIT: hmm, you found a bug... this bug can be fixed by adding the word auto_enable at the end of line 32 of share/scripts/lazy.tcl

By sd_snatcher

Prophet (3032)

sd_snatcher's picture

24-05-2014, 16:02

Don't ask me, all I know is that I upgraded openMSX to 0.10.1 and the reverse feature stopped working by default... Smile

By Manuel

Ascended (15631)

Manuel's picture

24-05-2014, 16:06

I hope the fix I mentioned helps, please check.

By sd_snatcher

Prophet (3032)

sd_snatcher's picture

24-05-2014, 16:58

Me and my unwanted gift for finding bugs... Smile

Here, the line 32 of lazy.tcl is:

register_lazy "_psg_log.tcl" psg_log

Since it didn't look to me as the correct line, I searched for "reverse", and found the following at line-39:

register_lazy "_reverse.tcl" {
        reverse_prev reverse_next goto_time_delta go_back_one_step
        go_forward_one_step reverse_bookmarks auto_save_replay_enable
        auto_save_replay_disable auto_save_replay_set_filename
        toggle_reversebar enable_reversebar disable_reversebar}

I added the auto_enable parameter just before the last brace and now the reverse function is working fine. Thanks!

BTW, I also found another bug. Have you seen the report?

Sorry if I wasn't able to test the prerelease of 0.10.1 when you asked in the forums, but I had no free time back then...

By Manuel

Ascended (15631)

Manuel's picture

24-05-2014, 21:26

Oops, 32 was a typo, but you found the right spot, thanks! I also put this in the current development version of course.

So, also flyguille will need this fix: https://sourceforge.net/p/openmsx/openmsx/ci/54bf3384e02d536...

The other report: I'm not good enough into the details of that, so I can't help there. I hope Wouter can respond.

By flyguille

Prophet (3029)

flyguille's picture

24-05-2014, 21:59

Manuel wrote:

How did you disable reverse in the first place then? It's on the 'gui' setting by default.

EDIT: hmm, you found a bug... this bug can be fixed by adding the word auto_enable at the end of line 32 of share/scripts/lazy.tcl

I thinks, instead of clicking in the window titlebar for seting back the keyboard focus on the MSX, I clicked like 2 pixel down off it. And it did the reverse. Then, maybe I panic and pressed some key with the arm or the left hand, and I never was capable of fowarding it again. Why not a confirmation box for reversing?.

By Manuel

Ascended (15631)

Manuel's picture

24-05-2014, 22:47

flyguille: because that's not very convenient when playing a game with reverse. Then you are doing reverse each time you make a mistake (e.g. with the PgUp shortcut) and directly overwrite the older 'future'.

As I said, perhaps you should just disable the reverse bar as I explained (although you'll need to make that little bug fix before the instructions will work).

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

25-05-2014, 23:07

why is this enabled in default, isn't it affecting performance?

By Manuel

Ascended (15631)

Manuel's picture

26-05-2014, 10:25

The reverse bar isn't affecting performance much. Just a tiny bit when it's visible.
The reverse feature itself mostly takes a bit more RAM (there are a certain number of savestates saved in RAM) and a little more CPU (to compress the savestates), but on modern systems you won't really notice. (On Dingoos and on Android, reverse is not enabled by default, because these systems tend to have a lot less RAM and CPU power.)

All in all, in general, it's more than worth the extra RAM/CPU power to be able to undo your mistakes. And if it's causing problems on a specific system, it can be easily disabled.

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

26-05-2014, 11:04

Ok, thanks for clearing that up. reverse stop in the console should suffice, I tried to save that setting in the console with save_settings , but didn't do anything, is this also due to that little bug?

By Manuel

Ascended (15631)

Manuel's picture

26-05-2014, 11:10

No, reverse stop is just a command to stop reverse, like reverse start is the command to start it.
However, there is that setting which I explained (auto_enable_reverse) which controls whether reverse is automatically started when starting openMSX. You need to modify that setting to change this behavior. If you never want reverse, just set it to off. The bug I mentioned only occurs when setting auto_enable_reverse to on (bug is: it doesn't go on).

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

26-05-2014, 11:45

ok, clear as water, thanks! OpenMSX blows BlueMSX out of the water btw great piece of software!

By Manuel

Ascended (15631)

Manuel's picture

26-05-2014, 12:18

Thanks for the compliments!

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

02-06-2014, 12:12

BTW, it did work out with the auto_enable_reverse off,
OpenMSX now starts wayy faster than before, it's instant, before with the option on, it did take on my HP dv7 at least 5 to 7 seconds for the graphical screen startupscreen to appear via Catapult. Now when I press start, it starts immediately. Smile

By Manuel

Ascended (15631)

Manuel's picture

02-06-2014, 14:25

Well, that's not normal at all. My 7 year old PC always starts it up instantly, and I always have reverse enabled.

Did you see any messages when it took so long to start up? (E.g. in the status tab of Catapult.)

By AxelF

Champion (340)

AxelF's picture

02-06-2014, 15:06

Will saving to .CAS files be possible in future versions ?
CAS files are easy to edit with use of a HEXeditor, and saves time
not have to convert WAV in to CAS every time.

Keep up the great work Smile

By wouter_

Champion (415)

wouter_'s picture

03-06-2014, 14:12

AxelF wrote:

Will saving to .CAS files be possible in future versions?

We currently don't have plans to add this.

Also it's not always possible to convert WAV to CAS, and you may only discover it wasn't possible after a lot of data has already been written. The best approach I can think of right now is to first buffer all data (like save to WAV) and then at the end try to convert it to CAS. But then you can almost as well use an external tool to convert WAV to CAS.

Maybe you have better ideas for how to reliably implement save to CAS? And of course you're always welcome to contribute a patch Smile

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

08-06-2014, 06:29

No, there were no messages in status tab. :S

But i found a bug, its not a constant, an occasional bug.

the situation is as following:
I have a floppy dsk file that I use to transfer files to my harddisk dsk file with openmsx, then in openmsx it compiles the files. I cannot leave my image filemanager open with m HD image and leave openmsx on, then it becomes read-only from one end, so the floppy image is my file buffer, that dsk I open firstly with my image filemanager, so it is read-only end is for openmsx. And then the whole session (few hours) i can leave these open and copy and compile all i want, but....

... with this i found out that speeding this up in openmsx also 'sometimes' goes wrong, cause the copying and compiling i do not use on 100% normal speed ofcourse. I put it on max with F9.

The bug is that with Max speed on, openmsx makes copy errors in my sourcefiles, extra characters which screw up so the compiler gives allot of errors or it puts the EOF marker somewhere else.

Then when i execute my batchfile (for copying and compiling) again, then all goes well.
I have this maybe 2 times out of 10.

By Manuel

Ascended (15631)

Manuel's picture

09-06-2014, 11:04

OK, I'm a bit confused.... so you are using a disk image and not dir-as-disk?
What kind of copying is going on? It copies from the disk image to the harddisk image?
Which openMSX version are you using exactly?

Can you provide me with the necessary files to reproduce this problem?

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

09-06-2014, 13:15

yes i am using disk image.

Copy is simple MSX-DOS2 copy from one small floppy DSK image to the other which is a 32mb HD DSK image, a few c source files are copied. I copy around 10 sourcefiles at once.(not all are compiled at once)
I am using Openmsx Version 0.10.1.
The errors occur during compiling in ASCII MSX-C 1.20. Cause the compiler then sees the errors in the files.

I first thought is was the compiler, but I opened the copied files with type and with TED and saw characters that not supposed to be there and arent in the originals.

I hope you have enough info. You can get some example files and compile them in this same kinda situation. But My guess is that this is during the copy stage, which is basically simple ascii text files.

here are some files, and also some complex sourcefiles to compile.
http://msx.hansotten.com/index.php?page=msx-c

By Manuel

Ascended (15631)

Manuel's picture

09-06-2014, 14:11

Can you confirm that the following test should be equivalent:

1) prepare a disk image with loads of ascii text files
2) start openMSX (which machine are you using exactly?) with IDE extension and the disk image (ends up in drive G: in my case)
3) in MSX-DOS do a COPY G:*.txt E: (I used E: as test destination) while full throttle (F9) is active
4) for ease of comparison: insert a dir-as-disk dir as diskb (drive HSmile, and do COPY E:*.txt H:
5) compare the txt files with the ones on the disk image of step 1

I tried this several times and each time the content was the same.

Can you try the same thing?

Also, can you reproduce the original issue and send me the original and modified files? (So I can see exactly what has changed in them.)

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

09-06-2014, 14:29

Will do tomorrow. The machine is a hb700p

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

10-06-2014, 11:41

Well, you always see when you really want them to happen, it doesnt hahah
I tried copying textfiles, they were without any error, i did 100x the copy/compile batch from dsk image, went perfectly and also the dir-as-dsk went perfectly 10x...
I am beginning to supsect my laptop, previous days i was in an very hot room, now i am in airconditioned room, otherwise i have noo idea. (i am on pre-honeymoon location, getting married on saturday)
If i do get it again today, i will report it, its not like i made this up ofcourse. -=confused=-

By Manuel

Ascended (15631)

Manuel's picture

11-06-2014, 17:55

OK, there's nothing I can do about it without more input from your side.

By Wolverine_nl

Paladin (1022)

Wolverine_nl's picture

11-06-2014, 22:10

I didnt do anything different today and yesterday then normal and in the cooled room everything seems to work fine, still happy with OpenMSX ;-)

Someone else thinks I had maybe dodgy code (which I changed already when trying to reduce bytes and simplify things) and that that caused it, but the compiler doesnt alter the source c files, and those weren't the same as the original.

Like I said, when it happens again, I will inform you right away and email you everything compressed, so you have the exact situation, now the project is further and the conditions are changed too much, so when it is in another stage it might happen again. I hope not Wink

By Manuel

Ascended (15631)

Manuel's picture

12-06-2014, 17:58

OK, thanks! And thanks for reporting in the first place!